Daspletosaurus sp.

Warner-Cowgill, Ethan, Storrs, Glenn W., Rogers, Raymond R. & Maltese, Anthony E., 2025, Cranial anatomy and stratigraphy of a new specimen of the tyrannosaurine dinosaur Daspletosaurus from the Judith River Formation of Central Montana, USA, Acta Palaeontologica Polonica 70 (1), pp. 159-174 : 163-169

publication ID

https://doi.org/10.4202/app.01143.2024

persistent identifier

https://treatment.plazi.org/id/03C10D5F-FFA3-0832-4654-FA625E1FFE35

treatment provided by

Felipe

scientific name

Daspletosaurus sp.
status

 

Daspletosaurus sp.

Figs. 4–10 View Fig View Fig View Fig View Fig View Fig View Fig View Fig .

Material.—CMC VP15826, partial skull and nearly complete postcranium from the Upper Cretaceous (Late Campanian) Coal Ridge Member of the Judith River Formation, Fergus County, Montana.

Description.—The specimen CMC VP15826, is approximately 80% complete, was partially articulated when discovered, and preserves portions of the skull and most of the vertebral column ( Fig. 4 View Fig ). The skull and cervical vertebrae were situated at the erosional edge of the specimen where preservation was poor. No tooth bearing elements other than a fragment of the left maxilla are present nor is the braincase or skull roof, but bones of the suspensoria and palate are preserved. A severely weathered portion of the posterior mandible may be present, although it is too poorly preserved to be described. The postcranial skeleton is very well-represented. Only three cervical vertebrae and a dozen or so caudal vertebrae are lacking, the latter primarily the distalmost elements. Most of the ribs and chevrons are present as is most, if not all, of the gastral basket. The pectoral girdle is complete with both scapulocoracoids, as well as the furcula. The forearms are similarly preserved, lacking only the ungual phalanges. Both ilia and ischia are preserved, but not the pubes. The left leg lacks only pedal phalanx IV-1 and most of metatarsal III; the right leg was not recovered.

CMC VP15826 is very large, with a reconstructed total body length greater than 9 meters. The ontogenetic maturity of this specimen is attested to not only by its size but also by the complete fusion of almost all neurocentral sutures down to the most distally preserved caudal vertebrae, fused scapulae and coracoids, and in particular the extreme rugosity of the bone surfaces of the neural spines. The distal ends of the spines, and especially the anterior and posterior surfaces of the spines are heavily textured with hyperostoses reflecting partial ossification of the spinous ligaments. Similar patterning is present at the distal ends of the transverse processes.

CMC VP15826 is assigned to Daspletosaurus by the presence of two cranial synapomorphies: (i) a deep keel on the ventral surface of the vomer and (ii) a posterior pneumatic recess of the palatine located posterior to the anterior margin of the vomeropterygoid neck ( Carr et al. 2017; Voris et al. 2020). CMC VP15826 further agrees with Daspletosaurus and differs from the daspletosaurin Thanatotheristes in having a mediolaterally thin suborbital margin of the jugal Voris et al. 2020). CMC VP15826 can be excluded from D. horneri ( Carr et al. 2017) by the presence of a maxillary fenestra that is anteriorly tapered, longer than tall, and dorsally offset from the ventral margin of the antorbital fossa, a ventral pneumatic recess of the squamosal that is not entirely undercut, the absence of a pneumatic foramen on the neck of the quadratojugal, the presence of a dorsal quadratojugal process of the jugal positioned lateral to the ventral quadratojugal process, an inflation of the jugal that extends below the pneumatic recess, and a humerus-femur length ratio of ~0.37 ( Carr et al. 2017). CMC VP15826 differs from D. wilsoni (sensu Warshaw and Fowler 2022; Warshaw et al. 2024) and agrees with D. torosus (CMN 8506) in having a medially directed dorsal quadrate process of the quadratojugal.

Because D. horneri and D. wilsoni have been primarily distinguished from D. torosus by their skulls, the present study focuses on, but is not limited to, cranial elements. A comprehensive description of the well-represented postcranial skeleton of CMC VP15826 is beyond the scope of this paper but will be the subject of a future study. A table of measurements of the postcranial skeleton is provided in SOM (Supplementary Online Material available at http:// app.pan.pl/SOM/app70-WarnerCowgill_etal _SOM.pdf).

A humerus-to-femur length ratio of ~0.34 has been considered autapomorphic of D. horneri ( Carr et al. 2017) . The holotype of D. torosus does not preserve a femur, so Carr et al. (2017) used the length of the skull and the ilium as a proxy and found ratios of 0.34 and 0.32, respectively. The femur of CMC VP15826 is 97 cm long as measured from the proximal surface to the medial condyle. The right humerus is 34 cm, whereas the left humerus is 38 cm, a difference that is probably a result of postmortem distortion. By averaging the length of both humeri, the humerus to femur length ratio in the present specimen is ~0.37, a ratio most similar to Gorgosaurus libratus as described by Lambe (1917), but much higher than Tyrannosaurus rex (0.29, Brochu 2003) and Tarbosaurus bataar (0.26, Carr et al. 2017). The ilia are 108 cm (right) and 111 cm (left) in length, providing a femur-to-ilium length ratio of 0.87, and a humerus-to-ilium length ratio of ~0.33. Carr et al. (2017) report that D. horneri has stout epipophyses that either just reach or only slightly project past the postzygapophyses, whereas D. torosus is coded in that study as having long epipophyses. Though incompletely preserved, the epipophyses on the anterior-mid cervicals of CMC VP15826 are short and do not reach the postzygapophyses, reflecting the reported condition in D. horneri . Based on unpublished photographs of the original specimens, we have independently verified the observations of Carr et al. (2017) regarding the morphology of the cervical epipophyses in D. horneri and the holotype of D. torosus . Additional specimens of D. torosus are, however, necessary to determine if the presence of long epipophyses is taxonomically informative or intraspecifically variable.

Maxilla: Although only a fragment of the left maxilla is present ( Fig. 5A View Fig ), important features are preserved. The fragment represents the area immediately surrounding the anterior part of the maxillary fenestra. As in other specimens of Daspletosaurus , the promaxillary fenestra is hidden from lateral view and the maxillary fenestra was large. Although the absolute size of the maxillary fenestra is not known because the interfenestral strut is not preserved, it was greater than 98 mm in length and greater than 78 mm in height. The maxillary fenestra is anteriorly tapered as in D. torosus and D. wilsoni (Warshaw and Fowler 2022) , unlike the round condition in D. horneri ( Carr et al. 2017) . As in D. torosus and D. wilsoni the maxillary fenestra of CMC VP15826 is dorsally offset from the ventral margin of the antorbital fossa whereas in subadult (MOR 590) and adult (MOR 1130) D. horneri the fenestra is positioned closer to the ventral margin of the antorbital fossa. The promaxillary fenestra measures ~ 13 mm in width and is ~ 43 mm long and has a ventral margin located slightly below the anterior corner of the maxillary fenestra. The presence of anastomosing sulci on the subcutaneous surface of the maxilla, a synapomorphy of Daspletosaurus ( Carr et al. 2017) , cannot be assessed in CMC VP15826 due to poor preservation.

Jugal: The left jugal ( Fig. 5B View Fig ) is nearly complete and ca. 480 mm long, but taphonomic distortion has left the distal portion of the postorbital ramus severely crushed, and the entire ramus bent posteriorly. This may be explained by the fact that the jugal was the only cranial element that was oriented vertically in the quarry during excavation. This element aligns with D. torosus rather than D. horneri in that the lateral inflation extends below the pneumatic recess, the ventral quadratojugal process is positioned medial to the dorsal quadratojugal process, and the lateral maxillary process is visible in medial view ( Carr et al. 2017).

As in D. torosus and D. wilsoni the ventral margin of the orbit is extremely thin mediolaterally as opposed to the thin but rounded margin of Lythronax argestes and the wide rounded margin in Thanatotheristes degrootorum (Voris et al. 2020; Warshaw and Fowler 2022). The posterior portion of the lacrimal contact is very shallowly inclined as in D. wilsoni and D. torosus (Warshaw and Fowler 2022) , whereas this surface is very steep in D. horneri , albertosaurines Currie 2003), and Lythronax argestes . Anterior to the orbit the jugal is directed straight anteriorly as in D. torosus and D. wilsoni (Warshaw and Fowler 2022) .

The jugal of CMC VP15826 differs from the holotypes of D. horneri , D. wilsoni , and D. torosus in a few ways. The suborbital height of the jugal is much more dorsoventrally constricted than in the holotypes of D. torosus , D. wilsoni , and D. horneri . However, the same is also true for TMP 2001.36.1, a complete skull from the Oldman Formation of Alberta provisionally identified as D. torosus by Voris et al. (2019: fig. 6B), a new taxon by Paulina-Carabajal et al. 2021), and most recently referred to D. wilsoni by Warshaw et al. (2024). The antorbital margin of the jugal in both TMP 2001.36.1 and CMC VP15826 also share a more shallowly inclined dorsal margin than in D. horneri (MOR 590 and MOR 1130), or the holotypes of D. torosus (CMN 8506) and

D. wilsoni (BDM 107). The proximal portion of the postorbital ramus of the jugal in both TMP 2001.36.1 and CMC VP15826 is relatively narrow anteroposteiorly, whereas in D. horneri, CMN 8506, and BDM 107 it is anteroposteriorly broad. Whereas D. torosus (CMN 8506) and D. wilsoni (BDM 107) have a large pneumatic opening on the maxillary ramus of the jugal (pneumatic recess, Fig. 5B View Fig ), in both CMC VP15826 and TMP 2001.36.1 they are reduced in size. The extreme reduction in size of the pneumatic recess in CMC VP15826 may be the result of an overgrowth of bone reflecting the advanced ontogenetic stage of this individual. A laterally protuberant accessory cornual process is present posterior to and above the primary ventral cornual process. This process is located well anterior to the quadratojugal contact and is therefore distinct from the knob that is autapomorphic of Teratophoneus curriei ( Carr et al. 2011) . This process is undercut and has a rounded, convex surface. Such an accessory cornual process has not been reported in other specimens of Daspletosaurus and may be pathological.

Squamosal: The left squamosal ( Fig. 6 View Fig ) is largely complete and 216 mm in maximum length, although the anteroventral edge of the quadratojugal process is missing. In the absence of the postorbitals, it cannot be positively stated that the squamosal terminates posterior to the anterior margin of the laterotemporal fenestra, a synapomorphy of Daspletosaurus ( Carr et al. 2017) . However, the squamc

mosal of CMC VP15826 does not significantly differ from D. torosus in any important respect. As in D. torosus and the holotype of D. wilsoni (Warshaw and Fowler 2022) , the anteromedial margin of the ventral pneumatic recess of the squamosal is not undercut, whereas this margin is autapomorphically undercut in D. horneri ( Carr et al. 2017) . A large foramen on the medial surface of the posterior process indicates that this process was pneumatic as in all species of Daspletosaurus .

Quadratojugal: Both quadratojugals ( Fig. 7 View Fig ) are preserved, but the right is complete and well preserved, whereas the left is lacking half of its anterior (jugal) ramus and the process forming the dorsal quadrate process. The maximum dorsoventral height of the right quadratojugal is 217 mm and the length along its ventral margin is 180 mm. The quadratojugals agree with the holotype of D. torosus and differ from the holotypes of D. horneri and D. wilsoni in having a posteromedially directed, rather than a posterolaterally directed dorsal quadrate process. Unlike D. horneri , the anterior ramus is oriented anteroventrally rather than anterodorsally. It agrees with D. torosus and D. wilsoni but differs from D. horneri (MOR 1130) in lacking a pneumatic foramen on the lateral surface of the neck and in having a deeply notched ventral quadrate process. CMC VP15826 shares one feature in common with D. horneri absent from the holotypes of D. torosus (CMN 8506) and D. wilsoni (BDM 107); the facet for the dorsal jugal process is partly obscured from lateral view whereas in CMN 8506 and BDM 107 this facet is entirely exposed in lateral view (Warshaw and Fowler 2022; Fig. 7A View Fig 2 View Fig ). The anterior margin of the anterior ramus of the quadratojugal is squared off, rather than tapered or forked.

Quadrate: Both left and right quadrates ( Fig. 7B View Fig ) are complete and relatively well preserved despite some transverse crushing. They average 242 mm in maximum dorsoventral height, 290 mm in maximum length through the pterygoid flange (“orbital process” of Carr et al. 2017), and 144 mm in maximum mediolateral width across the mandibular condyles. Among species of Daspletosaurus , only the quadrates of D. wilsoni have been adequately figured, limiting comparisons between specimens. The quadrates of CMC VP15826 do not appear to differ markedly from the holotype of D. wilsoni (Warshaw and Fowler 2022: fig. 9) but do differ from D. horneri in lacking fossae ventral to the anterior pneumatic recess, in lacking an undercut posteroventral margin of the medial fossa of the pterygoid flange, and for having a pneumatic recess that does not reach the medial surface of the pterygoid flange ( Carr et al. 2017).

The left and right quadrates of CMC VP15826 are asymmetric with respect to one another in several ways. In the right quadrate the dorsal surface of the pterygoid flange is straight for most of the length of the flange, whereas in the left quadrate this margin is deflected sharply ventrally at the anterior third of its length similar to the condition in D. wilsoni (BDM 107). A groove is reportedly present between the articular condyles of D. torosus , but not in D. horneri Carr et al. 2017 ). In CMC VP15826 there is a groove on the right quadrate only, with the condyles being divided instead on the left quadrate by a broad concavity. The appearance of a groove on the right quadrate may be a result of postmortem distortion, however. The angle of the anteroventral margin of the pterygoid flange of the left quadrate appears to be essentially identical to that of D. wilsoni , though this angle is reportedly steeper in D. horneri than in D. torosus Warshaw and Fowler 2022 ; Carr et al. 2017). In the present specimen, however, this margin is considerably less steep in the right quadrate. Such asymmetries in a single individual may indicate a high degree of plasticity in the morphology of this element.

Vomer: The fused vomers ( Fig. 8 View Fig ) are nearly complete anterior to the point where both the left and right sides of the element come together, and approximately 446 mm long as preserved. Though twisted and crushed, it is largely consistent with the morphology of other tyrannosaurids ( Brochu 2003; Currie 2003). The left and right sides together enclose an elongated groove on the posterior half of the dorsal surface of the element. The anterior plate is broad relative to the mediolateral width of the stem, but not to the extraordinary degree of lateral expansion present in Tyrannosaurus rex (Molnar 1991; Larson 2008). No foramina are present within the anterior plate as there are in some specimens of Tyrannosaurus rex ( Larson 2008) . As in Tyrannosaurus rex , there is a mediolaterally narrow, dorsoventrally flat process that continues anterior to the anterior plate and that forms the anterior limit of the element ( Brochu 2003), although it appears to be incomplete in the present specimen. A deep keel is present at the midline of the element’s ventral surface, a synapomorphy of Daspletosaurus sp. ( Carr et al. 2017). A keel is also present on the ventral side of the vomer in Tyrannosaurus rex below the posterior end of the anterior plate, whereas in Daspletosaurus sp. this keel is positioned well posterior to the plate ( Larson 2008).

Pterygoid: Both pterygoids are represented, but only the right element ( Fig. 9 View Fig ) is nearly complete and relatively well preserved. The right pterygoid is approximately 440 mm long and is similar in essence to that of T. rex and other tyrannosaurids ( Larson 2008). The element is mediolaterally thin for its entire length. Posteriorly it possesses a dorsally projecting quadrate process that overlaps the anterior part of the medial surface of the pterygoid flange of the quadrate. A prominent anterodorsally directed ramus of the pterygoid contacts the mediodorsal surface of the palatine, and together these elements enclose a rounded fenestra ( Figs. 9 View Fig and 10 View Fig ) that is roughly ~ 38 mm tall and ~ 30 mm long.

Ectopterygoid: The left ectopterygoid ( Fig. 10A View Fig ) is mostly complete, but much of its surface is badly crushed. The anterior or jugal ramus curls anteroventrally from the body of the element and bears a dorsoventrally broad, flattened surface for contacting the jugal. The ventral ramus curls posteroventrally from the body and has a single anteroposteriorly elongate (~ 40 mm) pneumatic excavation on its medial surface as in D. horneri , whereas two foramina are present here in the holotype of D. torosus ( Carr et al. 2017) . There is a ~ 10 mm wide accessory pneumatic foramen in the body of the ectopterygoid between the jugal and ventral processes.

Palatine: The main body of the right palatine ( Fig. 10B View Fig ) is largely intact, but most of the vomeropterygoid process as well as the distal ends of the jugal process, medial process, and maxillary ramus are missing. As preserved, the bone has an anteroposterior length along its ventral margin of approximately 245 mm; its maximum dorsoventral height is just over 183 mm. The posterior pneumatic recess is located behind the anterior margin of the vomeropterygoid neck, a synapomorphy of the genus Daspletosaurus ( Carr et al. 2017) . Posteromedially, there is a process for contacting the lateral surface of the anteroventral part of the pterygoid which, together with the posterior margin of the dorsal ramus of the palatine, forms a fenestra between these elements ( Figs. 9 View Fig and 10 View Fig ).

As the holotype of D. wilsoni does not include a palatine, CMC VP15826 is here compared only to D. horneri and D. torosus . The palatine agrees closely with the holotype of D. torosus . Features shared with the holotype of D. torosus but not D. horneri include closely spaced pneumatic recesses, an internal partition between the pneumatic recesses, a posterior pneumatic recess that is as tall as the maxillary ramus, and the presence of a medial pneumatic opening ( Carr et al. 2017). The vomeropterygoid neck of the palatine is oriented anterodorsally as in D. torosus (CMN 8506) rather than nearly vertically as in the holotype of D. horneri ( Carr et al. 2017) . The posterior pneumatic recess is rounded posteriorly and narrows anterodorsally as in D. torosus , whereas the recess in the paratype of D. horneri is not as tapered and has a more vertically oriented anterior margin Carr et al. 2017: fig 2F).

Darwin Core Archive (for parent article) View in SIBiLS Plain XML RDF